Overview
Syllabus
00:00 Introduction - Amjad Masad
00:42 Historical perspective
02:22 How AI can change software
04:29 ⚠️ Announcing AI for all!
06:33 A tale of Code LLM & GPU-Poor - Michele Catasta
07:29 How Replit's code completion works
08:39 ⚠️ Announcing Replit's new model!
13:45 ⚠️ Announcing the new model is open source!
14:06 Model training
15:26 Model evaluation
17:31 Model data & training
18:45 Model evaluation
19:51 Model inference
22:10 Why open source?
23:10: Glaive AI Collaboration
23:50 Morph Labs Collaboration
24:42: Perplexity Collaboration
Taught by
AI Engineer